Storage System and Method for Controlling Block Rearrangement

1. A storage system, interconnected by a first network with a plurality of host computers sharing the storage system, comprising: a plurality of disk array apparatuses, one of the plurality of disk array apparatuses having a disk array control device which controls a disk array includes a plurality of disk drives and which analyzes a file-level access request from one of the plurality of host computers to convert the file-level access request into an access request for one of the plurality of disk drives which correspond to the file-level access request; and a plurality of file system control devices interconnected by a second network with the plurality of disk array apparatuses, one of the plurality of file system control devices including a file system which manages a file specified in the access request from the one of the plurality of host computers, the one of the plurality of file system control devices receiving the access request from the one of the plurality of host computers to output the request to the one of the plurality of disk array apparatuses, the one of the plurality of file system control devices comprising: a fragmentation information collecting section which collects, from the file system, fragmentation information indicative of usage of blocks in a target volume for which a fragmentation state is to be eliminated, the target volume being included in those volumes in which the file managed by the file system is stored; an analysis section which analyzes the fragmentation state of the target volume on the basis of the fragmentation information collected by the fragmentation information collecting section, to generate block rearrangement information required to rearrange blocks in use in the target volume, in consecutive areas in the volume; and a block rearrangement command generating section which generates a block rearrangement command row required to cause the disk array control device to execute block rearrangement indicated by the block rearrangement information, the block rearrangement command generating section issuing the command row to the disk array control device, the disk array control device comprising: a block rearranging section which controls the disk array in accordance with the block rearrangement command row issued by the block rearrangement command generating section, to execute a block rearranging process of moving the data in the blocks in use in the target volume to the consecutive areas, and an access statistical information table which holds, for each of the logical units constituting the volume, access statistical information containing a logical address and a size of an area targeted by accesses, and a number of accesses per unit time to the area, the fragmentation information collecting section collects the access statistical information on the logical units constituting the target volume from the access statistical information table of the disk array control device when collecting fragmentation information on the target volume, the analysis section calculates, on the basis of results of analysis of the fragmentation state of the target volume and the access statistical information collected by the fragmentation information collecting section, a rate of access performance degradation caused by the fragmentation, for each area indicated by the access statistical information, from the number of accesses per unit time to the area and a number of physically inconsecutive areas into which the area is divided, determines priorities in block rearrangement for rearranging blocks in use in the area to be physically consecutive in accordance with the rate of access performance degradation, and generates the block rearrangement information with the priorities, the block rearrangement command generating section generates a block rearrangement command row with the priorities added to each command, from the block rearrangement information with the priorities, and the block rearranging section executes, in accordance with the corresponding priorities, a block rearranging process for each command on the basis of the block rearrangement command row.

2. A method of controlling block rearrangement, the method being used in a storage system which is interconnected by a first network with a plurality of host computers sharing the storage system, comprising a plurality of disk array apparatuses, one of the plurality of disk array apparatuses having a disk array control device which controls a disk array includes a plurality of disk drives and which analyzes a file-level access request from one of the plurality of host computers to convert the file-level access request into an access request for one of the plurality of disk drives which correspond to the file-level access request, and a plurality of file system control devices interconnected by a second network with the plurality of disk array apparatuses, one of the plurality of file system control devices including a file system which manages a file specified in the access request from the one of the plurality of host computers, the one of the plurality of file system control devices receiving the access request from the one of the plurality of host computers to output the request to the one of the plurality of disk array apparatuses, the disk array control device being used to execute block rearrangement required to eliminate a fragmentation state, the method comprising: collecting, from the file system, fragmentation information indicative of usage of blocks in a target volume for which the fragmentation state is to be eliminated, the target volume being included in those volumes in which the file managed by the file system is stored; collecting, from an access statistical information table which holds, for each of the logical units constituting the volume, access statistical information containing a logical address and a size of an area targeted by accesses, and a number of accesses per unit time to the area, the access statistical information on the logical units constituting the target volume from the access statistical information table of the disk array control device when collecting fragmentation information on the target volume; analyzing the fragmentation state of the target volume on the basis of the collected fragmentation information; calculating, on the basis of results of analysis of the fragmentation state of the target volume and the collected access statistical information, a rate of access performance degradation caused by the fragmentation, for each area indicated by the access statistical information, from the number of accesses per unit time to the area and a number of physically inconsecutive areas into which the area is divided, determining priorities in block rearrangement for rearranging blocks in use in the area to be physically consecutive in accordance with the rate of access performance degradation, and generating the block rearrangement information with the priorities; generating a block rearrangement command row required to cause the disk array control device to execute block rearrangement indicated by the generated block rearrangement information; and issuing the generated command row to the disk array control device to cause the disk array control device to execute a block rearranging process of moving the data in the blocks in use in the target volume to the consecutive areas.

3. A storage medium which stores program code applied to a storage system, which is interconnected by a first network with a plurality of host computers sharing the storage system, comprising a plurality of disk array apparatuses, one of the plurality of disk array apparatuses having a disk array control device which controls a disk array includes a plurality of disk drives and which analyzes a file-level access request from one of the plurality of host computers to convert the file-level access request into an access request for one of the plurality of disk drives which correspond to the file-level access request, the program code being used to execute processing on a plurality of file system control devices interconnected by a second network with the plurality of disk array apparatuses, one of the plurality of file system control devices including a file system which manages a file specified in the access request from the one of the plurality of host computers, the one of the plurality of file system control devices receiving the access request from the one of the plurality of host computers to output the request to the one of the plurality of disk array apparatuses, the program code, when executed, causing a processor to execute a method comprising: collecting, from the file system, fragmentation information indicative of usage of blocks in a target volume for which the fragmentation state is to be eliminated, the target volume being included in those volumes in which the file managed by the file system is stored; collecting, from an access statistical information table which holds, for each of the logical units constituting the volume, access statistical information containing a logical address and a size of an area targeted by accesses, and a number of accesses per unit time to the area, the access statistical information on the logical units constituting the target volume from the access statistical information table of the disk array control device when collecting fragmentation information on the target volume; analyzing the fragmentation state of the target volume on the basis of the fragmentation information collected; calculating, on the basis of results of analysis of the fragmentation state of the target volume and the collected access statistical information, a rate of access performance degradation caused by the fragmentation, for each area indicated by the access statistical information, from the number of accesses per unit time to the area and a number of physically inconsecutive areas into which the area is divided, determining priorities in block rearrangement for rearranging blocks in use in the area to be physically consecutive in accordance with the rate of access performance degradation, and generating the block rearrangement information with the priorities; and generating a block rearrangement command row required to cause the disk array control device to execute block rearrangement indicated by the block rearrangement information generated and issuing the command row to the disk array control device to cause the disk array control device to execute a block rearranging process of moving the data in the blocks in use in the target volume to the consecutive areas.